Revisiting Storage for Smartphones. Nitin Agrawal

Size: px

Start display at page:

Download "Revisiting Storage for Smartphones. Nitin Agrawal"

Garey George
6 years ago
Views:

1 Revisiting Storage for Smartphones Hyojun Kim Nitin Agrawal Cristian Ungureanu

Life in the Post-PC Mobile Era Smartphone and tablet markets are huge & growing 100 Million

(& growing) access through mobile; mobile users twice as active [FB] Mobile Innovation in

Samsung Galaxy Note Hardware add-ons: NEC Medias (6.

2 Life in the Post-PC Mobile Era Smartphone and tablet markets are huge & growing 100 Million smartphones shipped in Q4 2010, 92 M PCs [IDC] Out of 750 Million Facebook users, 250 Million (& growing) access through mobile; mobile users twice as active [FB] Mobile Innovation in mobile hardware: packing everything you need in your pocket Blurring the phone/tablet divide: Samsung Galaxy Note Hardware add-ons: NEC Medias (6.7mm thick, waterproof shell, TV tuner, NFC, HD camera,..) Manufacturers making it easier to replace PCs Motorola Atrix dock converts a phone into laptop 2

3 3

4 4

5 Waiting is undesirable! Annoying for the user More so for interactive mobile users Easy to lose customers More time, more battery Aren t network and CPU the real problem? Why are we talking about storage? 5

6 Understanding Mobile Performance Well understood! Network performance can impact user experience 3G often considered the bottleneck for apps like browsing Service providers heavily investing in 4G and beyond CPU and graphics performance crucial as well Plenty of gaming, video, flash-player apps hungry for compute Quad-core CPUs, GPUs to appear on mobile devices Not well understood! Does storage performance impact mobile experience? For storage, vendors & consumers mostly refer to capacity 6

7 Wireless Network Throughput Progression Standard (theoretical) Measured in Lab Mobile Flash a/g 3G Flash storage on mobile performs better than wireless networks Most apps are interactive; as long as performance exceeds that of the network, difficult for storage to be bottleneck 7

8 Outline Introduction Why storage is a problem Android storage background and setup Experimental results Solutions 8

9 Performance MB/s Why Storage is a Problem Random versus Sequential Disparity Performance for random I/O significantly worse than seq; inherent with flash storage Mobile flash storage classified into speed classes based on sequential throughput Random write performance is orders of magnitude worse Vendor (16GB) Speed Class Cost US $ Seq Write Rand Write Transcend RiData Sandisk Kingston Wintec A-Data Patriot PNY Consumer-grade SD performance However, we find that for several popular apps, substantial fraction of I/O is random writes (including web browsing!) 9

Why Storage is a Problem Shifting Performance Bottlenecks Standard (theoretical) Measured in Lab Mobile Flash Seq 802.

10 Why Storage is a Problem Shifting Performance Bottlenecks Standard (theoretical) Measured in Lab Mobile Flash Seq A/G 3G Mobile Flash Rand Storage coming under increasingly more scrutiny in mobile usage Random I/O performance has not kept pace with network improvements n (600 Mbps peak) and ad (7 Gbps peak) offer potential for significantly faster network connectivity to mobile devices in the future 10

11 Deconstructing Mobile App Performance Focus: understanding contribution of storage How does storage subsystem impact performance of popular and common applications on mobile devices? Performed analysis on Android for several popular apps Several interesting observations through measurements Storage adversely affects performance of even interactive apps, including ones not thought of as storage I/O intensive SD Speed Class not necessarily indicative of app performance Higher total CPU consumption for same activity when using slower storage; points to potential problems with OS or apps Improving storage stack to improve mobile experience 11

12 Outline Introduction Why storage is a problem Android storage background and setup Experimental results Solutions 12

Storage Partitions on Android /misc 896KB settings /recovery rootfs 4MB alternate boot /boot rootfs 3.

3MB read write /sdcard FAT32 16GB read write Internal NAND Flash Memory (512MB) External SD Partition Misc Recovery

boot-into-recovery partition for advanced recovery Enables the phone to boot, includes the bootloader and kernel

system images Stores user data (e.g., contacts, messages, settings) and installed apps; SQLite DB containing app data also stored here.

13 Storage Partitions on Android /misc 896KB settings /recovery rootfs 4MB alternate boot /boot rootfs 3.5MB kernel /system yaffs2 145MB read-only /cache yaffs2 95MB read write /data yaffs MB read write /sdcard FAT32 16GB read write Internal NAND Flash Memory (512MB) External SD Partition Misc Recovery Boot System Cache Data Sdcard Sd-ext Function H/W settings, persistent shared space between OS & bootloader Alternative boot-into-recovery partition for advanced recovery Enables the phone to boot, includes the bootloader and kernel Contains the remaining OS, pre-installed system apps ; read-only Used to stage and apply over the air updates; holds system images Stores user data (e.g., contacts, messages, settings) and installed apps; SQLite DB containing app data also stored here. Wiped on factory reset External SD card partition to store media, documents, backup files etc Non-standard partition on SD card that can act as data partition 13

14 Phone and Generic Experimental Setup Rooted and set up a Google Nexus One phone for development GSM phone with a 1 GHz Qualcomm QSD8250 Snapdragon processor 512 MB RAM, and 512 MB internal flash storage Setup dedicated wireless access point b/g on a laptop for WiFi experiments Installed AOSP (Android Open Source Project) Linux kernel modified to provide resource usage information 14

15 Custom Experimental Setup Requirements beyond stock Android Ability to compare app performance on different storage devices Several apps heavily use the internal non-removable storage To observe and measure all I/O activity, we modified Android s init process to mount all internal partitions on SD card Measurement study over the internal flash memory and 8 external SD cards, chosen 2 each from the different SD speed classes Observe effects of shifting bottlenecks w/ faster wireless networks But, faster wireless networks not available on the phones of today Reverse Tethering to emulate faster networks: lets the smartphone access the host computer s internet connection through a wired link (miniusb cable) Instrumentation to measure CPU, storage, memory, n/w utilization Setup not typical but allows running what-if scenarios with storage devices and networks of different performance characteristics 15

Using HTTP proxy server App Install Top 10

YouTube GasBuddy, Gmail, Twitter, Books,

benchmark Facebook Android Email Google Maps

16 Apps and Experiments Performed WebBench Browser Visits 50 websites Based on WebKit Using HTTP proxy server App Install Top 10 apps on Market App Launch Games, Weather, YouTube GasBuddy, Gmail, Twitter, Books, Gallery, IMDB RLBench SQLite Synthetic SQL benchmark Facebook Android Google Maps Pulse News Reader Background Apps: Twitter, Books, Gmail Contacts, Picasa, Calendar Widgets: Pulse, YouTube, News, Weather, Calendar, Facebook, Market, Twitter 16

17 Outline Introduction Why storage is a problem Android storage background and setup Experimental results (talk focuses on runtime of apps) Paper has results on I/O activity, CPU, App Launch behavior, etc Solutions 17

Time (seconds) Time (s) WebBench Results: Runtime 4000 3000 2000 1000 0 WIFI USB 100 80 60 40 20 0 WiFi USB Time taken for iperf to download 100MB Runtime on WiFi varies by 2000% between internal and

18 Time (seconds) Time (s) WebBench Results: Runtime WIFI USB WiFi USB Time taken for iperf to download 100MB Runtime on WiFi varies by 2000% between internal and Kingston Even with repeated experiments, with new cards across speed classes Even without considering Kingston, significant performance variation (~200%) Storage significantly affects app performance and consequently user experience With a faster network (USB in RT), variance was 222% (without Kingston) With 10X increase in N/W speed, hardly any difference in runtime 18

Time (seconds) Time (s) WebBench Results: Runtime 600 500 400 300 200 100 0 WIFI USB 100 80 60 40 20 0 WiFi USB Time taken for iperf to download 100MB Runtime on WiFi varies by 2000% between internal

19 Time (seconds) Time (s) WebBench Results: Runtime WIFI USB WiFi USB Time taken for iperf to download 100MB Runtime on WiFi varies by 2000% between internal and Kingston Even with repeated experiments, with new cards across speed classes Even without considering Kingston, significant performance variation (~200%) Storage significantly affects app performance and consequently user experience With a faster network (USB in RT), variance was 222% (without Kingston) With 10X increase in N/W speed, hardly any difference in runtime 19

Runtimes for Popular Apps (without Kingston) 200 FaceBook 200 Maps 40 Email 100 100 20 0 0 0 400 App Install 200 RLBench 100 Pulse News 200 100 50 0 0 We find a similar trend for several popular

20 Runtimes for Popular Apps (without Kingston) 200 FaceBook 200 Maps App Install 200 RLBench 100 Pulse News We find a similar trend for several popular apps Storage device performance important, better card faster apps Apart from the benefits provided by selecting a good flash device, are there additional opportunities for improvement in storage? 0 20

21 WebBench: Sequential versus Random I/O I/O Breakdown Vendor Seq:Rand perf ratio Rand IOPS Transcend Sandisk RiData Kingston Wintec A-Data Patriot PNY Few reads, mostly at the start; significantly more writes About 2X more sequential writes than random writes Since rand is worse than seq by >> 2X, random dominates Apps write enough randomly to cause severe performance drop Paper has a table on I/O activity for other apps 21

How Apps Use Storage? Exactly what makes web browsing slow on Android? Key lies in understanding how apps use SQLite and FS interface /data/data/com.

webview WebBench Storage Schema lib (empty) cache webviewcache These files written to FS in write-behind These files written to SQLite in sync

22 How Apps Use Storage? Exactly what makes web browsing slow on Android? Key lies in understanding how apps use SQLite and FS interface /data/data/com.necla.webview WebBench Storage Schema lib (empty) cache webviewcache These files written to FS in write-behind These files written to SQLite in sync databases webview.db (14KB) 6aaa3f00, 03051d8d, many files (5.5MB) webviewcache.db (129KB) Apps typically store some data in FS (e.g., cache files) and some in a SQLite database (e.g., cache map) All data through SQLite is written synchronously slow! Apps often use SQLite oblivious to performance effects 22

Time (seconds) What-If Analysis for Solutions What is the potential for improvements? E.g., if all data could be kept in RAM? Analysis to answer hypothetical questions A. Web Cache in RAM B.

23 Time (seconds) What-If Analysis for Solutions What is the potential for improvements? E.g., if all data could be kept in RAM? Analysis to answer hypothetical questions A. Web Cache in RAM B. DB (SQLite) in RAM C. All in RAM D. All on SD w/ no-sync Baseline Placing Cache on Ramdisk does not WebBench on RiData improve perf. much DB on Ramdisk Both Cache and DB alone improves on SD without sync perf. significantly recoups most perf Both Cache and DB in RAM no extra benefit A B C D Cache in RAM DB in RAM All in RAM Disable fsync 23

24 Implications of Experimental Analysis Storage stack affects mobile application performance Depends on random v/s sequential I/O performance Key bottleneck is ``wimpy storage on mobile devices Performance can be much worse than laptops, desktops Storage on mobile being used for desktop-like workloads Android exacerbates poor storage performance through synchronous SQLite interface Apps use SQLite for functionality, not always needing reliability SQLite write traffic is quite random further slowdown! Apps use Android interfaces oblivious to performance Browser writes cache map to SQLite; slows cache writes a lot 24

25 Outline Introduction Why storage is a problem Android storage background and setup Experimental results Solutions 25

Time (seconds) Time (seconds) Pilot Solutions RAID-0 over SD card and internal flash Leverage I/O parallelism already existent Simple software RAID driver with striped I/O As expected speedup, along

26 Time (seconds) Time (seconds) Pilot Solutions RAID-0 over SD card and internal flash Leverage I/O parallelism already existent Simple software RAID driver with striped I/O As expected speedup, along with super linear speedup due to flash idiosyncrasies (in paper) Back to log-structured file systems Using NilFS2 to store SQLite databases Moderate benefit; suboptimal implementation Application-specific selective sync Turn off sync for files that are deemed async per our analysis (e.g., WebCache Map DB) Benefits depend on app semantics & structure PCM write buffer for flash cards Store performance sensitive I/O (SQLite DB) Small amount of PCM goes a long way Base RAID PCM RAM LogFS SelSync Base RAID PCM RAM LogFS SelSync WebBench on RiData

27 Conclusion Contrary to conventional wisdom, storage does affect mobile application performance Effects are pronounced for a variety of interactive apps! Pilot solutions hint at performance improvements Small degree of application awareness leads to efficient solutions Pave the way for robust, deployable solutions in the future Storage subsystem on mobile devices needs a fresh look We have taken the first steps, plenty of exciting research ahead! E.g., poor storage can consume excessive CPU; potential to improve energy consumption through better storage 27

28 We are hiring! Storage Systems Group

Understanding Storage I/O Behaviors of Mobile Applications. Louisiana State University Department of Computer Science and Engineering

Understanding Storage I/O Behaviors of Mobile Applications Jace Courville jcourv@csc.lsu.edu Feng Chen fchen@csc.lsu.edu Louisiana State University Department of Computer Science and Engineering The Rise